During the past decade the prevalence of ecstasy use has increased in the U.S. and in other countries. particularly among adolescents and young adults. Potential determinants that predict future ecstasy use in adolescence are understudied. Specifically, there is a lack of longitudinal research designed to control temporal sequencing and estimate the extent to which factors predict increased risk of ecstasy use. To address these gaps, the proposed study aims to: 1) test if the prevalence and incidence of ecstasy use in adolescence vary in relation to demographic and geographic variables over time; 2) test whether previous deviant behaviors, previous high sensation-seeking scores and prior other drug use predict subsequent ecstasy use among adolescents; 3) test whether previous low levels of parent monitoring and high levels of peer ecstasy use predict subsequent ecstasy use among adolescents; 4) test for changes in attitudes and beliefs towards ecstasy use among adolescents and test whether exposure to an anti-drug media campaign might have influenced adolescents' attitudes towards ecstasy use. Longitudinal data from the National Survey of Parents and Youth (NSPY), an epidemiologic study representative of the adolescent U.S. population (9-18 yeas old) will be used to address these aims. After initial exploratory data analyses, we will analyze data through cross-sectional and longitudinal logistic regression models to address each of the specific aims. First, we will estimate associations of prevalence and incidence of ecstasy use with demographics (age, gender, race/ethnicity), then, we will use alternating logistic regression to test whether there is evidence of clustering of adolescent ecstasy use by region of the country. Using longitudinal regression marginal models, we will test whether previous deviant behaviors, high sensation-seeking scores and high levels of prior other drug use predict subsequent ecstasy use among adolescents, as well as examine if lower levels of parent monitoring and high levels of peer ecstasy use predict subsequent ecstasy use among adolescents. Logistic regression models will test for changes in attitudes and beliefs towards ecstasy use among adolescents and to test whether exposure to an anti-drug media campaign influenced adolescents' attitudes towards ecstasy use. Finally, we believe that our study will bring public health contributions to better understand potential determinants of ecstasy use. [unreadable] [unreadable]